U.S. patent number 8,729,486 [Application Number 13/135,122] was granted by the patent office on 2014-05-20 for modfet active pixel x-ray detector.
This patent grant is currently assigned to The Board of Trustees of the Leland Stanford Junior University. The grantee listed for this patent is Henry M. Daghighian, Craig S. Levin, Peter D. Olcott, Farhad Taghibakhsh. Invention is credited to Henry M. Daghighian, Craig S. Levin, Peter D. Olcott, Farhad Taghibakhsh.
United States Patent |
8,729,486 |
Daghighian , et al. |
May 20, 2014 |
MODFET active pixel X-ray detector
Abstract
Detection of ionizing radiation with modulation doped field
effect transistors (MODFETs) is provided. There are two effects
which can occur, separately or together. The first effect is a
direct effect of ionizing radiation on the mobility of electrons in
the 2-D electron gas (2DEG) of the MODFET. An ionizing radiation
absorption event in or near the MODFET channel can perturb the 2DEG
mobility to cause a measurable effect on the device conductance.
The second effect is accumulation of charge generated by ionizing
radiation on a buried gate of a MODFET. The conductance of the
MODFET can be made sensitive to this accumulated charge, thereby
providing detection of ionizing radiation. 1-D or 2-D arrays of
MODFET detectors can be employed to provide greater detection area
and/or spatial resolution of absorption events. Such detectors or
detector pixels can be integrated with electronics, such as
front-end amplification circuitry.
Inventors: |
Daghighian; Henry M. (Santa
Clara, CA), Olcott; Peter D. (Stanford, CA), Levin; Craig
S. (Palo Alto, CA), Taghibakhsh; Farhad (Redwood City,
CA) |
Applicant: |
Name |
City |
State |
Country |
Type |
Daghighian; Henry M.
Olcott; Peter D.
Levin; Craig S.
Taghibakhsh; Farhad |
Santa Clara
Stanford
Palo Alto
Redwood City |
CA
CA
CA
CA |
US
US
US
US |
|
|
Assignee: |
The Board of Trustees of the Leland
Stanford Junior University (Palo Alto, CA)
|
Family
ID: |
45525763 |
Appl.
No.: |
13/135,122 |
Filed: |
June 23, 2011 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20120025087 A1 |
Feb 2, 2012 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
61398351 |
Jun 23, 2010 |
|
|
|
|
61519334 |
May 19, 2011 |
|
|
|
|
Current U.S.
Class: |
250/370.1 |
Current CPC
Class: |
H01L
29/66462 (20130101); H01L 31/119 (20130101); G01T
1/24 (20130101); G01T 1/00 (20130101) |
Current International
Class: |
G01T
1/24 (20060101) |
Field of
Search: |
;250/370.1 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
Other References
Romero et al., "An analytical model for the photodetection
mechanisms in high-electron mobility transistors", 1996, pp.
2279-2287, IEEE Trans. microwave theory and techniques, v44n12.
cited by applicant .
Claspy et al., "High frequency GaAlAs modulator and photodetector
for phased array antenna applications", 1988, pp. 1-12, NASA
technical memorandum 101328. cited by applicant .
Boardman et al., "Design and characterization of high electron
mobility transistors for use in a monolithic GaAs X-ray imaging
sensor", 2001, pp. 226-231, Nuclear instruments and methods in
plasma research, A 466. cited by applicant .
Hofstetter et al., "Development and evaluation of gallium
nitride-based thin films for x-ray dosimetry", May 4, 2011, pp.
3215-3231, Physics in medicine and biology, v56. cited by applicant
.
Lee et al., "Nitride-based MSM photodetectors with a HEMT structure
and a low-temperature AlGaN intermediate layer", 2008, pp.
H959-H963, Journal of hte electrochemical society, v155n12. cited
by applicant .
Fetterman et al., "Integrated optically driven millimeter wave
sources and receivers", 1994, pp. 1493-1496, IEEE MTT-S Digest.
cited by applicant .
Hofstetter et al., "Real-time x-ray response of biocompatible
solution gate AlGaN/GaN high electron mobility transistor devices",
Mar. 5, 2010, pp. 092110 1-092110 3, Applied Physics Letters, v96.
cited by applicant.
|
Primary Examiner: Taningco; Marcus
Attorney, Agent or Firm: Lumen Patent Firm
Parent Case Text
CROSS REFERENCE TO RELATED APPLICATIONS
This application claims the benefit of U.S. provisional patent
application 61/398,351, filed on Jun. 23, 2010, entitled "High
speed radiation detection using MODFET semiconductor device", and
hereby incorporated by reference in its entirety. This application
also claims the benefit of U.S. provisional patent application
61/519,334, filed on May 19, 2011, entitled "MODFET active pixel
X-ray detector", and hereby incorporated by reference in its
entirety.
Claims
The invention claimed is:
1. Apparatus for high speed detection of ionizing radiation, the
apparatus comprising: a modulation doped field effect transistor
(MODFET) having a channel for current flow between a source and a
drain, wherein electrons in the channel form a two-dimensional
electron gas (2DEG), and wherein the MODFET includes a first gate
to set an operating point of the MODFET for radiation detection;
and an absorption region disposed at and/or in proximity to the
channel and capable of absorbing ionizing radiation; wherein the
2DEG provide a conductance that is modulated in response to
absorption of ionizing radiation in the absorption region; wherein
the ionizing radiation has a photon or particle energy of about 20
keV or greater.
2. The apparatus of claim 1, wherein the electrons in the 2DEG have
a mobility that is modulated in response to absorption of ionizing
radiation in the absorption region.
3. The apparatus of claim 1, wherein the MODFET is a
double-heterostructure MODFET, and further comprising a buried
second gate disposed in proximity to the channel, wherein charges
generated by absorption of ionizing radiation in the absorption
region can accumulate on the buried second gate as accumulated
charge, and wherein the conductance is modulated in response to the
accumulated charge.
4. The apparatus of claim 1, wherein conductance changes caused by
incident ionizing radiation are measured via their effect on
operation of the MODFET.
5. The apparatus of claim 1, wherein the MODFET and absorption
region are of different material systems.
6. A sensor array comprising two or more of the apparatus of claim
1 disposed in a 1-D or 2-D array of sensor pixels.
7. The apparatus of claim 6, wherein the array of sensor pixels is
read out in parallel or sequentially.
8. An integrated circuit comprising the apparatus of claim 1.
9. The apparatus of claim 1, wherein the MODFET is implemented in
layers of Hg(1-x)Cd(x)Te that are grown lattice-matched to a layer
of Cd(1-y)Zn(y)Te that serves as the absorption region.
10. The apparatus of claim 1, wherein the conductance decreases in
response to absorption of ionizing radiation in the absorption
region.
11. A method for detection of ionizing radiation, the method
comprising: providing a modulation doped field effect transistor
(MODFET) having a channel for current flow between a source and a
drain, wherein electrons in the channel form a two-dimensional
electron gas (2DEG), and wherein the MODFET includes a first gate
to set an operating point of the MODFET for radiation detection;
and disposing an absorption region at and/or in proximity to the
channel that is capable of absorbing ionizing radiation; wherein
the electrons in the 2DEG provide a conductance that is modulated
in response to absorption of ionizing radiation in the absorption
region; wherein the ionizing radiation has a photon or particle
energy of about 20 keV or greater.
12. The method of claim 11, wherein the electrons in the 2DEG have
a mobility that is modulated in response to absorption of ionizing
radiation in the absorption region.
13. The method of claim 11, wherein the MODFET is a
double-heterostructure MODFET, wherein the MODFET further comprises
a buried second gate disposed in proximity to the channel, wherein
charges generated by absorption of ionizing radiation in the
absorption region can accumulate on the buried second gate as
accumulated charge, and wherein the conductance is modulated in
response to the accumulated charge.
14. The method of claim 11, further comprising measuring
conductance changes caused by incident ionizing radiation via their
effect on operation of the MODFET.
15. The method of claim 11, further comprising pulse width
modulation of detector signals to provide a modulated output
signal, wherein pulse width in the output signal is related to one
or more incident ionizing radiation parameters.
16. A method of single photon detection of ionizing radiation
comprising performing the method of claim 11 for an incident
photon.
17. The method of claim 16, wherein variation of a drain-source
current in the MODFET is indicative of absorption of the incident
photon, and the extent of the variation is indicative of the energy
of the incident photon.
18. The method of claim 17, further comprising passing a digital
data signal through the MODFET, wherein distortion of the digital
data signal is indicative of absorption of the incident photon, and
the extent of the distortion is indicative of the energy of the
incident photon.
19. The method of claim 11, wherein the conductance decreases in
response to absorption of ionizing radiation in the absorption
region.
Description
FIELD OF THE INVENTION
This invention relates to detection of ionizing radiation.
BACKGROUND
Semiconductor devices have been employed for some time in
connection with detection of ionizing radiation. Often, the charge
liberated by a detection event is directly measured. An article by
Boardman et al., (Nuclear Instruments and Methods in Physics
Research A 466 (2001) 226-231) is representative of this approach.
Here, a semi-insulating GaAs wafer serves as a detector element and
as the substrate for an integrated charge readout matrix. Charges
generated in this substrate are collected by the readout matrix to
provide the sensor output.
Another well known approach is the use of the photoconductive
effect, where incident ionizing radiation generates charge carriers
in a photoconductive device, thereby increasing its
conductance.
However, it remains challenging to provide detection of ionizing
radiation with detectors that simultaneously provide high
performance (e.g., high speed and high sensitivity) and ease of
integration with other circuitry, such as front end amplification
electronics. Accordingly, it would be an advance in the art to
provide such sensors for ionizing radiation.
SUMMARY
The mobility of electrons in the channel of a field effect
transistor (FET) can be greatly improved by removing scattering
centers (e.g., ionized donors) from the channel. Transistors that
exploit this principle are referred to as modulation doped FETs
(MODFETs) or high electron mobility transistors (HEMTs). Current
flow between source and drain in such transistors flows through a
channel which includes a 2-D electron gas formed in undoped
material. The high mobility of the electrons in this 2-D electron
gas results from the relative lack of electron scattering centers
in undoped material, compared to doped material, and results in
high speed and high amplification power of MODFET or HEMT
transistors.
In the present invention, these characteristics of MODFET/HEMT
devices are exploited to provide detection of ionizing radiation
(e.g., X-rays, gamma rays, and/or particles such as electrons and
positrons). More specifically, if ionizing radiation is absorbed by
the undoped material of a MODFET in the vicinity of the channel,
physical effects such as impact ionization, plasma renormalization
and/or donor ionization can lead to the formation of temporary
electron scattering centers. These scattering centers can affect
the mobility of electrons in the channel, and thereby provide a
sensor output responsive to received ionizing radiation.
Alternatively, a buried gate in MODFETs based on double
heterostructures can be employed to collect charges generated by
absorption of ionizing radiation. Charge on the buried gate can
strongly affect the channel conductance of the MODFET, thereby
providing alternative methods of detecting ionizing radiation. 2-D
arrays of MODFET detectors can be employed to provide greater
detection area and/or spatial resolution of detection events.
These approaches provide significant advantages. It is possible to
provide high speed photon counting for ionizing radiation of
sufficiently high energy, and it further provides the possibility
of high speed energy measurement for each detected ionizing
radiation photon. This approach is applicable for improving
performance (e.g., improving signal to noise ratio (SNR)) in any
imaging application, such as digital mammography, X-ray imaging,
computed tomography, etc. Improved SNR can directly lead to other
advantages, such as reduced radiation dose to a patient for the
same image quality. Photon counting+energy analysis cannot be
performed with present-day X-ray detector technology.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows a simple structure of a generic MODFET (prior
art).
FIG. 2 shows the energy band diagram relating to the MODFET
structure of FIG. 1 (prior art).
FIG. 3 is an exemplary device structure for a double-MODFET (prior
art).
FIG. 4 shows the energy band diagram relating to the double-MODFET
structure of FIG. 3 (prior art).
FIG. 5 shows an embodiment of the invention exploiting a MODFET as
a high speed radiation detector.
FIG. 6 shows an embodiment of the invention having a buried gate in
a double-heterostructure MODFET.
FIG. 7 shows calculated band structures and carrier concentrations
for the example of FIG. 6.
FIG. 8 shows an exemplary approach for performing radiation
measurements using sensors according to embodiments of the
invention.
FIG. 9 shows an example of a suitable phase frequency detector for
use in the example of FIG. 8.
FIG. 10 shows an integrated circuit including sensor elements
according to embodiments of the invention.
DETAILED DESCRIPTION
A) Basic MODFET Operation
Modulation doped field effect transistors (MODFETs) operate based
on the presence of an ultra-fast electronic conduction channel,
often referred to as a 2-dimensional electron gas (2DEG). FIG. 1
shows an example of a MODFET device, and FIG. 2 shows a
corresponding energy band diagram. In this example, a conductance
between a source 102 and a drain 104 is controlled by a gate 106. A
barrier layer 108 separates gate 106 from channel layer 110, built
on the substrate (112). Barrier layer 108 and channel layer 110
form a heterostructure as shown on FIG. 2. Barrier layer 108 has a
higher bandgap than channel layer 110. Channel layer 110 is
un-doped, while barrier layer 108 is doped. The doping of layer 108
is preferably away from layer 110 (e.g., as can be obtained with
the delta doping technique, or the inclusion of doped and undoped
sub-layers within barrier layer 110). The Fermi level is shown on
FIG. 2 as a horizontal line 215. Adjusting the voltage at gate 106
shifts the Fermi level vertically on FIG. 2, thereby affecting the
electron concentration in the channel layer. More specifically, a
2-D electron gas (2DEG, 114) in the quantum well 214 can form at
the interface between layers 108 and 110.
The hetero-junction formed by different band gap materials forms a
quantum well (i.e., a triangular well having a knife edge) in the
conduction band of the un-doped channel layer. Electrons in the
2DEG are confined to the quantum well, and therefore do not
experience scattering from ionized donors because the ionized
donors are away from the quantum well. The effect of this is to
create a very thin layer of highly mobile conducting electrons with
high concentration giving the channel very high electron mobility.
Thus, such transistors are also known as high electron mobility
transistors (HEMTs). In this description, the terms HEMT and MODFET
are regarded as synonyms and used interchangeably.
The 2-D electron gas in MODFETs allows super fast electrical
current transmission across the source and drain of the MODFET and
it is available largely due to lack of charged ions and coulomb
scattering centers in the un-doped channel layer. 600 GHZ bandwidth
has been achieved and MMIC (microwave monolithic integrated
circuit) structures are commonplace with HEMT technology.
A1) MODFET Types
In a pseudomorphic HEMT (PHEMT) a lower noise/faster MODFET
structure can be achieved by using an extremely thin layer of one
of the materials (i.e., so thin that the crystal lattice of the
thin layer stretches to fit the other material). This technique
allows the construction of transistors with larger band gap
differences than otherwise possible, giving them better
performance.
Another way to use materials of different lattice constants is to
place a buffer layer between them. This is done in the MHEMT or
meta morphic HEMT (e.g., using InAs and AlInAs). In this example,
the buffer layer is made of AlInAs with indium concentration graded
so that it can match the lattice constant of both the substrate and
the n-type channel. This brings the advantage that practically any
indium concentration of the channel can be realized so that devices
can be optimized for different applications (e.g., low indium
concentration provides low noise, and high indium concentration
provides high gain).
A2) MODFET: Basic Operation as an Amplifier
By applying a reverse bias voltage to the gate of a MODFET, the
voltage applied to the gate alters the conductance of the channel
by altering the carrier concentration in the 2DEG. As more negative
bias is applied, the 2d electron gas is depleted. This results in
the modulation of the channel conductance. Gain and amplification
occur until the channel is pinched off (fully depleted). The
transconductance is given by g.sub.m={acute over
(.epsilon.)}v.sub.satw.sub.g/d where {acute over (.epsilon.)} is
the permittivity, v.sub.sat is the saturated velocity, w.sub.g is
the unit gate width of device, and d is the distance from the gate
to the 2d electron gas. Since conduction of electrons from source
to drain occurs in a channel that is well confined, g.sub.m will be
kept high with low drain currents. A3) Specific MODFET Examples
(Single-MODFETs)
MODFET technology is well known in the art, so the MODFET structure
shown on FIG. 1 is simplified for convenience of exposition.
Single-MODFET structures have been used in variety of
configurations for detection and sensing of chemicals (Ren: US
patent application 2011/0068372), magnetic field (Folks: US patent
application 2008/0088982), or even ionizing radiation. Hofstetter
et al (in Applied Physics Letters, 96, 2010) report on application
of such a device for radiation detection where the radiation is
absorbed in a liquid medium and resulting charges are coupled to
the gate of the device. In all these, and other similar works, the
generated charge as the result of detection is coupled to the gate
of the single-MODFET device to modulate the channel conductance. In
an integrated approach, Boardman (Nuclear Instruments and Methods
in Physics Research A 466-2001) reports on use of MODFETs as a
passive switch for conducting the radiation induced charges to a
readout circuit (a passive pixel architecture).
A4) Specific MODFET Example (Double-MODFETs)
MODFET technology is also well known for high power applications,
where the high mobility channel helps reduce the power loss in the
device and prevent over heating. In order to further improve the
channel conductivity for power applications, double
heterostructures haven been used in MODFETs resulting in devices
generally known as double-MODFETs or double-HEMTs. In
double-MODFETs the configuration of layers around the channel is
symmetric in order to widen the channel for better conductance.
FIG. 3 shows an example of a double-MODFET structure. Here 102, 104
and 106 are the source, drain and gate, as above. Layer 302 is n+
GaAs, layer 304 is n-type AlGaAs, layer 306 is undoped AlGaAs (a
first spacer layer), 308 is an n+ delta doping within 306, layer
310 is undoped GaAs (the channel layer as above), layer 306' is
undoped AlGaAs (the second spacer layer), 308' is n+ delta doping
within 306', layer 312 is n-type AlGaAs, layer 314 is p GaAs
(buffer layer) and layer 316 is a GaAs semi-insulating
substrate.
As shown in the band diagram of FIG. 4, the double heterostructure
of FIG. 3 results in a wide quantum well (formed by layer 310) for
better channel conduction compared to the knife-edge narrow quantum
well band structure of a single-MODFET (e.g., as shown on FIG.
2).
B) MODFET Based Fast Radiation Detection
MODFET transistors (and their relatives such as double-MODFETs,
PHEMT, MHEMT etc.) rely on an un-doped substrate with no ionized
donors/coulomb scattering centers to maintain the 2-D electron gas
operation. In presence of high energy radiation such as X-rays and
gamma rays and nuclear particles, the substrate material of a
MODFET can experience impact ionization and/or plasma
renormalization. This is shown schematically as absorption 522 of
ionizing radiation 520 on FIG. 5. High-Z substrate materials are
preferred because of their greater absorbance of ionizing
radiation. The energy transfer from the ionizing radiation results
in perturbation of the 2D electron gas (e.g., by creating temporary
coulomb scattering centers that act similar to ionized donor
scattering centers). As a result, the conductance of the channel
between the MODFET source and drain can be affected by the ionizing
radiation, thereby enabling the MODFET to act as a radiation
sensor.
An example method of the present invention is shown in FIG. 5 where
a MODFET (single or double) is used for detection and measurement
of radiation absorbed in an absorbing region 112 which is disposed
at and/or in proximity to channel 110, and is capable of absorbing
ionizing radiation. As used herein, ionizing radiation refers to
electromagnetic or particle radiation having a photon or particle
energy of about 20 keV or greater. Absorption 522 of ionizing
radiation 520 in absorbing region 112 can alter the conductance
provided by the electrons in the 2DEG, thereby enabling the MODFET
to act as a sensor for ionizing radiation. In many cases, it is
preferred for absorbing region 112 to be fabricated in a different
material system than the other MODFET layers (e.g., 102, 104, 108,
110).
C) Buried Gate Double-MODFET
In a preferred embodiment, a buried gate is added to the
double-MODFET configuration as described above to further enhance
radiation detection. FIG. 6 shows an example of this approach.
Here, the conductance of a channel for current flow between a
source 102 and a drain 104 is set by a gate 106. As described
above, a 2DEG 114 can form in channel layer 310 to provide a path
for this current flow. Layers 108 and 310 form a heterostructure,
as do layers 310 and 112, thereby making the device of this example
a double-heterostructure MODFET. The device of this example can
directly read charge created in, or injected into, bulk substrate
(i.e., absorption region) 112 using a buried n-doped region that
operates as a gate 616 underneath channel layer 310. More
specifically, absorption 522 of ionizing radiation 520 can generate
charge which can accumulate on buried gate 616 (as schematically
shown by 624). It is convenient to refer to this kind of device as
a double-gate double-MODFET (DGDMODFET) device. In some cases, it
can be useful to make contact to the back side of the device with
back side contact 618.
It is worth noting that unlike in conventional field effect
transistors (such as silicon MOS transistors) where a buried gate
can be simply added below the channel (for example to form a
depFET), a buried gate will not work if added to a conventional
MODFET; buried gates work on double-MODFETs only. Double
heterostructures are essential for proper operation of such
devices.
Practice of the invention does not depend critically on the
materials used for the MODFET sensor. Any materials capable of
absorbing ionizing radiation and compatible with MODFET technology
can be employed. Suitable material systems include, but are not
limited to: SiGe, GaAs, HgCdTe, and CdZnTe. In one preferred
embodiment, the MODFET is fabricated in the HgCdTe system and is
lattice matched to the absorption region which is grown in the
CdZnTe system.
C1) Buried Gate MODFET Theory
To better appreciate the buried gate MODEFT approach, it is helpful
to consider a simple model of device performance. We define the
following quantities:
C.sub.g is effective capacitance per unit area of the un-doped
channel between the gate and the channel.
.mu..sub.n is the mobility of electrons in the 2DEG.
V.sub.gs is the gate source voltage between the back gate and the
channel.
V.sub.t is the sum of the intrinsic threshold voltage of the device
plus the top metal gate-source voltage (from gate 106).
.DELTA.Q is the charge collected at the back gate 116 from the
substrate 112.
C.sub.bg is the sum of all capacitance that the buried gate
sees.
The basic device equations are given by:
.times..times..mu..function. ##EQU00001##
.differential..differential..times..times..times..mu..function.
##EQU00001.2## .differential..DELTA..times..times. ##EQU00001.3##
.times. ##EQU00001.4##
The HEMT device has a higher mobility, and lower capacitance than
the standard MOSFET. Assuming the device is biased where V.sub.t is
constant and substituting in V.sub.gs, then the basic equations
give:
.DELTA..times..times..times..times..times..times..mu..function..DELTA..ti-
mes..times. ##EQU00002## Because the substrate is very lightly
doped, and the drain and source are relatively far away, we can
assume that C-parasitic is small compared to the gate capacitance,
which gives:
.DELTA..times..times..times..mu..times..DELTA..times..times.
##EQU00003##
Therefore, the change in current seen at the drain is inversely
related to the length of the device squared times the mobility of
the carriers in the 2DEG. Because the carriers in a HEMT device
travel in an un-doped layer, they can reach the theoretical
mobility of the semiconductor. Additionally, this device can
outperform MOS devices in terms of input referred noise because of
its higher mobility/gain and substantially smaller sensing node
capacitance (no oxide, no line or contact capacitance).
The charges on the buried gate can be cleared in two ways. First,
the buried gate can be engineered to have a certain carrier
lifetime by allowing the electrons to tunnel into the channel based
on the distance. Secondly, the top gate 106 of the device can be
forward biased, thereby clearing the accumulated charge from buried
gate 116.
FIG. 7 shows calculated band diagrams and carrier concentrations
for a buried gate MODFET (as in FIG. 6) in the GaAs/AlGaAs material
system. Two 1-D band diagrams for the heterostructure are taken at
the gate (line 630 on FIG. 6) and adjacent to the gate (line 640 on
FIG. 6). The buried gate is referenced as 616 on FIG. 7. The bands
labeled Ev1 and Ec1 are the bands along line 630 (through the
gates), and the bands labeled Ec2 and Ev2 are the bands along line
640 (away from the gates). Buried gate 616 will collect electrons
generated in, or injected into the substrate 112 in response to
ionizing radiation. Note the steep slope of Ec1 tending to force
electrons to the buried gate. The collection of charge at the
buried gate reduces the conduction in the channel 310, thereby
modulating the trans-conductance of the device. The conduction away
from the gates is higher (the top gate 106 can be negatively biased
to reduce channel conductance at the buried gate), so the device
conductance is strongly affected by charge on the buried gate.
Here, the gate terminal 106 is used for biasing the device at a
proper operating point as opposed to be conventionally used as the
sensing terminal. Alternatively 106 can be used to electrically
gate the device for synchronization. The accumulated charge on the
buried gate can be cleared by applying a proper voltage to the gate
106. This clearing mechanism would not be available in a
corresponding MOSFET based structure.
D) Buried Gate MODFET Sensors for Detection of Ionizing
Radiation
As indicated above, double-MODFETs having a buried gate are
sensitive to ionizing radiation because charge generated by
ionizing radiation accumulates on the buried gate, where it
effectively modulates the device conductance.
Thus, ionizing radiation can affect the buried gate MODFET
conductance via accumulation of charge on a buried gate, or via an
effect on the mobility of electrons in the 2DEG. These effects may
also occurs simultaneously in practice. Often the conductance will
decrease in response to accumulation of charge and/or reduction of
2DEG mobility, however it may also be possible for device
conductance to increase in response to ionizing radiation.
There are several approaches for sensing this conductance change.
As the energy of the ionizing radiation increases, it is expected
that the perturbation of the 2DEG in response to detection of a
single high energy event will be more prolonged. Thus, duration of
the perturbation is expected to provide energy resolution
capability in each of the following specific methods.
C1) Bit Error Rate Detection
By passing a stream of fast digital data bits through the HEMT
device and monitoring via a D-type flip flop phase frequency
detector (or comparator) the timing and levels, one can use this
device for detection of very short lived or rare events. Circuits
such as bit error rate detectors, PLL phase frequency detectors,
and comparators can all be combined with MODFET/HEMT devices to
form an integrated detector system.
As data is transmitted through the device, any perturbation to the
2-D electron gas as a result of X-ray radiation and gamma ray
energy transfer in the absorption region (e.g., via impact
ionization) will cause the bit transmission to suffer phase delay
and/or amplitude reduction. By monitoring or comparing the digital
bit transmission through the device we can identify radiation
events in <psec levels. For applications which involve Gamma
radiation (510 KeV), we can stack one or more high Z bulk absorbing
regions in proximity to MODFET detector channels (e.g., using wafer
bonding techniques or other conventional means) to provide a path
to slow down the gamma rays and allow energy transfer into the
MODFET detector substrate.
FIG. 8 schematically shows an example of this approach. A reference
data source 804 provides digital data to MODFET 82-1 (and
optionally to additional MODFETS 82-2 . . . 82-n, for an n-element
MODFET sensor array). MODFETs 82-1 . . . 82-n can each include
their own front end electronics for amplification. Optionally,
further amplification can be provided by a gain stage 806. The
MODFET outputs and reference data are provided to a phase frequency
detector 808, and the output of phase frequency detector 808 is
provided to a processor 810. FIG. 9 shows an exemplary circuit for
phase-frequency detector 808 on FIG. 8. Briefly, phase-frequency
detector 808 compares the data as passed through the MODFET sensors
with the original data, and processor 810 accumulates the results.
This bit error rate testing can be performed at high speeds (e.g.,
10 GHz, 25 GHz or even higher), which can provide high time
resolution for single-event detection.
This bit error test configuration is not only useful in terms of
picking up energy of X-ray photon but also reduces background noise
since we look for a specific signal pattern defined with bit rate
and bit sequence. It is in a way analogous to use of lock in
amplifiers and light choppers for weak optical signal
spectroscopy.
C2) Pulse Width Modulation
MODFET sensor technology as described herein is also compatible
with the pulse width modulation (PWM) approach of US patent
publication 2010/0025589, hereby incorporated by reference in its
entirety. Briefly, in the PWM approach, time and energy of a
detection event can be encoded into a single analog signal
pulse.
C3) Pulse Height Comparator
Data that is passed through a MODFET sensor can have its pulse
height monitored using a comparator (which can be integrated with
the MODFET sensor). Ionizing radiation is expected to perturb the
pulse height because of effect of ionizing radiation on the channel
conductance.
D) Examples
Commercially available, room temperature, nuclear spectra-graphic
grade, bulk grown, semiconductor crystals that can detect X-rays
include silicon, germanium, gallium-arsenide, cadmium-telluride,
and cadmium.sub.1-x-zinc.sub.x-telluride (x=0.1, 0.2).
The above-described MODFET detector structure can be grown on a
spectra-graphic grade absorbing bulk (or inexpensive liquid phase
grown) substrate. The available high Z, X-ray spectrographic grade,
bulk grown, materials include GaAs, CdTe, and
Cd.sub.0.9Zn.sub.0.1Te. Presently, there is no commercial supplier
of spectrographic wide-gap Hg.sub.0.2Cd.sub.0.8Te or
Hg.sub.0.3Cd.sub.0.7Te that is epitaxially or bulk grown.
Below are four examples of the above-described HEMT structure for
directly sensing, amplifying, and/or reading out X-rays for CT
applications.
D1) HEMT Readout Structure for Si
A silicon-germanium HEMT structure as described above can be grown
on bulk spectrographic grade silicon substrates. There are
commercial applications for the bulk silicon technology, but bulk
Si has not previously been employed for x-ray CT. To increase the
stopping power, the silicon substrate can be wafer bonded to higher
Z spectrographic grade materials such as CdZnTe.
D2) HEMT Readout Structure for GaAs
GaAs does not have the stopping power needed for efficient X-ray
CT, but the technology is extremely mature for building HEMT
structures. The above-described device can be manufactured by high
volume commercially available GaAs fabrication facilities. The
theory of device operation can be verified on GaAs easily.
D3) HEMT Readout Structure for Cd.sub.0.9Zn.sub.0.1Te
There exist MOVCD or MBE grown epitaxial Hg.sub.1-xCd.sub.xTe
(x=0.1-0.2) that is lattice matched for Cd.sub.0.9Zn.sub.0.1Te. It
is also known as MerCadTel and is a commercially available infrared
detector technology. It is possible to grow these
Hg.sub.1-xCd.sub.xTe HEMT structures directly on CdZnTe. This is
the current commercial production path used for infrared detectors.
Even though the device has a very low bandgap (Eg=0.3-0.5 eV),
because it so thin (1-2 um for the entire device), it can be
operated at room temperature. This technology has the promise to
greatly increase the count-rate performance of CdZnTe.
D4) HEMT Readout Structure for Thick Liquid Phase Epitaxially Grown
Widegap Hg.sub.0.3Cd.sub.0.7Te
Because of the very high Z of mercury (80), only 1.5 mm thickness
is needed to match the stopping power of 3 mm of CdZnTe. This range
is in the upper range of possible liquid-phase epitaxial growth of
Hg.sub.0.3Cd.sub.0.7Te on CdZnTe lattice matched substrates.
E) Advantages
E1) High Spatial Resolution Provided by Buried Gate Pixels
The buried gate MODFET configuration provides substantial
advantages for radiation detection. Because of the availability of
spectroscopic quality CdZnTe, high quality HgCdTe n-DGDHEMT
transistors can be grown on top of CZT substrates to provide
superior sensors. Doped buried gate regions grown in the thin
layers of HgCdTe can define the electrode pattern. The device is
very sensitive to the charge that collects on the gate, but is
relatively insensitive to the bulk leakage of the substrate. In
this way, the device effectively provides a small pixel area
similar to what can be obtained with metal electrodes with steering
grids. When grown on the substrate, the back contact (e.g., 118 on
FIG. 5) of the device can be the cathode contact of the CdZnTe
wafer.
E2) On-Chip Integration of Sensors with Front End Electronics
When comparing the HEMT device to an off the chip sensor readout
silicon device, it is apparent that a sensor pixel connected to an
external front end readout device will have a very large parasitic
capacitance because of the connection of the pixel to the front-end
readout device. A significant advantage of the present approach is
that this parasitic capacitance can be avoided by monolithic
integration of MODFET sensor elements with front-end
electronics.
The buried gate capacitance C.sub.bg can be on the order of 10-30
femto-F, which is significantly smaller than capacitances provided
by any hybrid technologies based on CdZnTe interconnected to
silicon readout chips. The small capacitances of the sensing
structures are a major advantage of active pixel sensors. Combined
with the extremely high speed of conduction in the channel, the
HEMT structure can drive large amounts of current at GHz speeds
into a sensing electrode. These currents then can then drive
off-chip circuitry, after receiving significant on-chip gain at
very high bandwidths.
E3) Radiation Hardness
The HEMT structures are more radiation hard than silicon MOSFET
devices because of the lack of a gate-metal-oxide and therefore do
not suffer from threshold voltage shifts formed by deep traps in
the MOSFET gate oxide. Many studies have been carried out for Co-60
irradiation (a good test of X-ray hardness) of the structure, and
shown device functionality with 10e6 Gray (1e9/rad) in GaAs and
silicon.
E4) 1-D and 2-D Sensor Arrays, and Integration
The present approach is compatible with the use of 1-D and/or 2-D
MODFET sensor arrays. Furthermore, MODFET detector elements can be
monolithically integrated with any other kind of circuitry (e.g.,
radiation-hard MODFET electronics). This capability advantageously
provides significant flexibility. For example, front end amplifier
electronics can be integrated with MODFET sensor elements. As
another example, wireless communication circuits can be integrated
with MODFET sensors. Combining these two ideas can provide a
wireless MODFET sensor head with integrated front end electronics.
MODFET sensors can be integrated with pattern generation and clock
distribution circuitry, which is especially beneficial for the bit
error rate detection approach. FIG. 10 schematically shows an
example, where an integrated circuit 1002 includes a 1-D MODFET
sensor array 1006 and/or a 2-D MODFET sensor array 1008 in addition
to processing and/or front end amplification circuitry 1004. Such
1-D or 2-D sensor arrays can be read out in parallel or
sequentially. Multiplexing techniques can be useful for sensor
arrays having a large number of elements.
* * * * *